How Does GA4 Work? From Data Collection to Reports

Google Analytics 4 (GA4) is an event-based analytics platform designed to track user interactions across websites and mobile applications. It replaces the session-centric model of Universal Analytics with a unified schema focused on events, parameters, and user properties.

What Is Google Analytics 4?

Google Analytics 4 is Google's primary analytics tool for measuring traffic, user engagement, and revenue. Unlike legacy analytics, GA4 uses a single measurement model for both web and app data, allowing cross-platform tracking within a single property.

How Does GA4 Work?

GA4 operates through a continuous three-stage pipeline: Collection, Processing, and Reporting.

Collection: Web and app triggers send event payloads to Google servers.

1. Processing: Google applies attribution models, resolves user identities, and organizes event parameters.

2. Reporting: Processed data populates real-time views, standard reports, custom explorations, and connected data warehouses.

Step 1: Create an Account, Property, and Data Stream

Setting up GA4 requires three hierarchical levels:

• Account: The top-level access management folder for your organization.

• Property: The container that processes and stores your analytics data.

• Data Stream: The data pipe connecting a specific source to your property.

To create a Data Stream:

1. Go to Admin > Data Collection and Modification > Data Streams.

2. Click Add stream and select Web, Android app, or iOS app.

3. Enter your website URL or app details, then click Create stream.

Step 2: Collect Data with the Google Tag, GTM, or an SDK

Data collection relies on code snippets deployed on your digital assets via the Google Tag (gtag.js), Google Tag Manager (GTM), or the Firebase SDK for mobile apps.

Automatically Collected Events

GA4 tracks basic actions automatically once the base code is installed.

• session_start

• first_visit

• user_engagement

Enhanced Measurement Events

Activate automated web tracking options without custom code under Admin > Data Streams > [Select Stream] > Enhanced Measurement.

• Page views (page_view)

• Scrolls (scroll)

• Outbound clicks (click)

• Site search (view_search_results)

• Video engagement (video_start, video_progress, video_complete)

• File downloads (file_download)

Recommended and Custom Events

• Recommended Events: Predefined names and parameters for standard industry verticals (e.g., purchase, generate_lead, add_to_cart).

• Custom Events: Unique event names and parameters created when a required business metric is not covered by default options.

Step 3: Attach Event Parameters, User Properties, and Identifiers

GA4 enriches raw events using context layers:

• Event Parameters: Key-value pairs describing event details (e.g., item_name, page_location, value).

• User Properties: Attributes defining the user persona (e.g., membership_tier, preferred_language).

• User Identifiers: Methods used to unify cross-device journeys:

  • User ID: First-party IDs assigned by your login system.

  • Google Signals: Data from logged-in Google users who consented to ads personalization.

  • Device ID: Browser cookies (client_id) or app instance IDs.

Step 4: Process Sessions, Users, and Engagement

During processing, GA4 aggregates raw hits into structured metrics.

• Users: Tracked via identity spaces to deduplicate visits across devices.

• Sessions: Initiated by session_start. Sessions end after 30 minutes of inactivity without a hard midnight reset.

• Engagement & Active Users: GA4 shifts focus from bounce rate to engagement metrics. See our detailed guides on Active Users in GA4 and GA4 Engagement Rate for in-depth calculations.

Step 5: Store and Aggregate Data

Processed data is saved in aggregate tables for quick reporting and underlying raw data structures for advanced queries.

Step 6: Explore Data in Realtime, Reports, and Explorations

Analyze data across three primary interfaces inside GA4:

• Realtime: Live monitoring of activity happening within the last 30 minutes (Reports > Realtime).

• Standard Reports: Out-of-the-box summary cards for acquisition, engagement, monetization, and retention (Reports > Acquisition).

• Explorations: Advanced workspace for custom funnels, path analysis, and freeform tables (Explore).

How Do Key Events and Attribution Work?

• Key Events: High-value actions marked in Admin > Data Display > Key Events to measure success metrics across campaigns.

• Attribution: Rules determining how credit for conversions is assigned to touchpoints. GA4 uses Data-Driven Attribution by default, distributing credit based on historical algorithmic models.

How Does GA4 Connect with Google Ads and BigQuery?

• Google Ads Link: Export GA4 Key Events for ad bidding and import Google Ads cost data into GA4 reports (Admin > Product Links > Google Ads Links).

• BigQuery Integration: Export raw, unaggregated event logs directly to Google Cloud for custom SQL analysis (Admin > Product Links > BigQuery Links).

What Privacy and Consent Controls Does GA4 Use?

• No IP Storage: GA4 does not log or store individual IP addresses.

• Consent Mode: Dynamically adapts tag behavior based on user cookie consent choices.

• Regional Data Controls: Disable granular location and device data collection per region.

How Do You Test a GA4 Implementation?

Use this checklist to verify your tracking setup:

• Tag Assistant: Open tagassistant.google.com to inspect outgoing hits from your site.

• GA4 DebugView: Enable debug mode in GTM or via the GA4 Debugger Chrome extension, then monitor events live under Admin > Data Display > DebugView.

• Realtime Report: Check Reports > Realtime to confirm events are registering under standard traffic logs.

Common GA4 Limitations and Reporting Differences

• Thresholding: Reports may hide small user counts to protect privacy when Google Signals is enabled.

• Cardinality: High-volume unique custom parameters can consolidate long-tail data into an (other) row.

• Non-Retroactive Data: Configuration changes (e.g., custom dimensions, Key Events) apply only from the time of creation forward.

Frequently Asked Questions (FAQ)

1. Is GA4 completely sessionless?

No. GA4 uses an event-based model, but it still automatically collects session_start events and measures sessions for user journey context.

2. Why do user counts in GA4 differ from Universal Analytics?

GA4 focuses on Active Users (users with an engaged session or active interaction) in standard reports, whereas UA emphasized Total Users.

3. Do I need BigQuery to use GA4?

No. BigQuery is optional, but linking it is recommended to bypass standard interface limits and retain long-term raw event data.